In my last blog post, I discussed some of the unexpectedly difficult things about being a professor. One thing I brought up was that it can be hard to get all of your advisees to give you feedback because of the advisor-advisee power dynamic. After I published the post, Yisong Yue shared his “Group Assimilation Exercise” that aims to encourage his students to give him collective feedback. I recently had my students undertake a similar exercise and was really happy with the result, so I'm sharing the process in hopes that it will be useful for anyone who wants to get honest collective feedback from their advisees.

The exercise

On the whole, I followed Yisong's approach pretty closely. The basic idea behind the exercise is this: You write down a list of questions to prompt discussion, then have your advisees meet without you to collectively discuss, brainstorm, and write down answers. Afterward, they share the feedback with you and you prepare some responses and action items to discuss with them. Why have your students meet as a group without you? First, if you directly ask someone who “reports” to you for feedback, they might not feel totally comfortable being brutally honest. Meeting without you helps ensure they feel they can speak openly. Second, meeting as a group makes it easier to brainstorm and validate each other's experiences.

Thinking of questions was not too hard for me since I spend a lot of time thinking about what's going well and not-so-well in my group. I also took inspiration from the list of questions that Yisong shared. Ultimately, I divided my questions into four categories: Research directions, lab culture, lab logistics, and advising. In each category, I included a bunch of questions that aimed to prompt discussion (rather than being an exhaustive list of the exact things they should speak about). I specifically asked them to brainstorm things that were going well, things that were not going well, and specific changes they'd suggest. Each category had ten or so questions; here are a few examples:

• Research directions - Are you happy with the overall research directions/topics we pursue in the group? Do you think we are working on impactful work? Are you excited about the work you/we are doing?
• Lab culture - Do you feel that there is a good sense of community in our group? Is our lab collaborative enough? Do you feel like you can be your real self among your labmates?
• Lab logistics - How do you feel about our lab space? Are the lab meetings helpful/interesting?
• Advising - Do you feel I have time to give you the help and advising you need? Am I doing enough to help promote your work/help you succeed in your career? Am I helpful in identifying areas for you to improve and helping you improve in them?

I wrote my questions, as well as an explanation of the exercise and what I hoped to get out of it, in a Google Doc and shared it with my students. If you are curious to see the full doc, it's here. Then, I designated one of our normal 2-hour lab meeting sessions as a time to do the exercise. I also ordered drinks and snacks from a local coffee shop to be delivered during the meeting as a small “thank you” for undertaking the exercise.

The results

Overall, I received a ton of useful feedback from doing this exercise and am really grateful that my students took it seriously and had a lot to say. Some of the feedback was simple and could be easily addressed with a specific action item; other feedback was higher-level and served more as an opportunity to discuss different perspectives. Here are a few examples of easy-to-address feedback:

• My students told me that after explaining something I often say “do you get what I'm saying?” but that can make the listener feel pressured to agree that they understand so that they seem knowledgeable. They suggested instead saying “did I explain that well?”. This makes a ton of sense and I immediately started doing it!
• During the pandemic, we had virtual gather town co-working sessions as a proxy for having a normal physical lab space (where coincidental co-working might spontaneously happen). We stopped doing that when our lab finally became in-person, but my students lamented the loss of a specific time to do co-working. So, going forward we will be scheduling specific office hours as times to encourage people to come in at the same time and work together.
• In various places, my students requested more regular high-level advising meetings (e.g. long-term research goals, retrospectives on how things are going, etc.). I had done these on an ad-hoc basis but will be doing them at the start of each semester going forward.
• There were various ideas for lab meeting topics (e.g. skill sharing, lightning talk sessions, etc.). In the past, I came up with all lab meeting topics myself, so this was a really nice opportunity to get other ideas and also remind my students they could suggest topics too.

The higher-level feedback was also valuable, not because it led to specific action items but because it provided an opportunity to explain various choices I had implicitly made. For example, my students shared that they weren't sure whether our research agenda was risky or broad enough, which led me to explain how I have tried to choose an agenda that balances exploration and exploitation while de-risking at the individual level. Since there's no way to know what the “correct” approach is here (especially in the short term), I also clarified that I expect my students to sometimes deviate from my approach/vision. While it was great to get to hear this feedback and discuss it, a downside was that I felt the discussion was probably more honest/vigorous when I wasn't there, but I probably needed to be in the loop for discussion of these topics in order to explain certain choices/approaches I've made. I'm considering trying to write down some of these approaches and sharing them with my students so they can use that as a point of reference for future feedback.

On the whole, I felt the exercise was extremely valuable and plan to do it yearly with my students. If you advise a group of people and feel you'd benefit from getting collective feedback, I'd recommend doing a similar exercise. Thanks again to Yisong for pointing it out to me and for providing the blueprint.